Release Time:2019-03-11  Hits:

Indexed by: Conference Paper

Date of Publication: 2010-01-01

Included Journals: CPCI-S

Page Number: 5670-5675

Key Words: fuzzy model; robustness; fuzzy clustering; system identification; H(infinity) estimation

Abstract: The paper proposes a new fuzzy identification method based on H(infinity) error estimation for the issues of robust identification of fuzzy model. The H(infinity) state estimation is applied to the parameter identification of fuzzy model in the paper. The presented algorithm not only guarantees to satisfy a specified level of robustness, and also provides an optimized error upper bound. Finally, we study the fuzzy model of nonlinear system. With the comparison between fuzzy identification based on recursive least square and the proposed algorithm in the paper, the simulated results show that the improving robustness of identification needs to be at the cost of approximation accuracy of identification.